Article Text

Download PDFPDF

Changing trends over the last decade in the aetiology of childhood blindness: a study from a tertiary referral centre
  1. Taylan Ozturk,
  2. Duygu Er,
  3. Aylin Yaman,
  4. A Tulin Berk
  1. Paediatric Ophthalmology and Strabismus Unit, Department of Ophthalmology, Dokuz Eylul University School of Medicine, Izmir, Turkey
  1. Correspondence to Professor A Tulin Berk, Paediatric Ophthalmology and Strabismus Unit, Department of Ophthalmology, Dokuz Eylul University School of Medicine, Akarsu Sok, Mesa 2.Etap, No. 5, Urla, Izmir 35430, Turkey; tulin.berk{at}, tulinberk61{at}


Aims To discern treatable and preventable causes of childhood blindness by evaluating the aetiologic factors, and to compare the distribution of the most commonly affected anatomic sites of severe visual impairment (SVI) with our previous published data.

Methods The charts of 11 871 patients followed between June 2002 and May 2014 were reviewed retrospectively, and 695 patients (5.9%) who had SVI or blindness in accordance with WHO criteria were enrolled. The results of ophthalmologic examinations and coexistence of any systemic disease were documented and checked against our published clinic data concerning the aetiology of childhood blindness before 2002. χ2 test was used for statistics.

Results Mean age was 47.0±51.9 months (median: 24 months). Cortical visual impairment (CVI) was present in 212 cases (30.5%) and 20.3% of those had a history of premature birth. The most common anatomic sites of SVI were retina (24.6%) and crystalline lens (17.1%). When compared with our previous data, we found a significant increase in the prevalence of CVI (p=0.046) and decrease in the frequency of SVI due to uveal disorders (p<0.001). Prevalence of blindness secondary to retinopathy of prematurity reduced by a third (p=0.280), and a significant decrease in aphakia-related SVI (p=0.028) was achieved within the last decade.

Conclusions The prevalence of CVI was found to be relatively increased due to the significant reduction in the frequency of preventable causes of SVI. Furthermore our clinical practice for visual rehabilitation in aphakia has resulted in a considerable decrease in SVI in the last decade.

  • Child health (paediatrics)
  • Vision

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Children who are born blind or become blind in later life not only suffer but also become a burden to their family and society for a lifetime. Since severe visual impairment (SVI) in children may affect their development, education and employment opportunities; early diagnosis is crucial for early intervention that might minimise the functional impact of blindness over society.1 ,2 Many causes of childhood blindness such as retinoblastoma, vitamin A deficiency, and metabolic syndromes may be related to childhood mortality or morbidity, so the early diagnosis of ocular disorders will affect the survival of the child.2–7

For the purpose of designing efficient programmes to prevent childhood blindness, obtaining reliable current data on major causes of SVI is mandatory. Herein, most common aetiologic factors and affected anatomic sites that could be related to childhood blindness were presented to discern treatable and preventable causes. Furthermore, comparison of our current results with previously published data from our tertiary referral centre aimed to highlight the trending topics in the aetiology of childhood blindness in Izmir, Turkey.8

Materials and methods

We retrospectively reviewed the charts of 11 871 patients seen between June 2002 and May 2014 in our Paediatric Ophthalmology and Strabismus Unit, which is a tertiary referral clinic and a part of a multispeciality university hospital. All patients underwent a full ophthalmologic examination, including fundoscopy, orthoptic assessment, cycloplegic refraction, and binocular single vision, where cooperation was adequate. Best-corrected visual acuity (BCVA) could be assessed using Snellen chart and Teller acuity cards whereas the optokinetic nystagmus and CSM (central, steady, maintained) were taken into account according to the bilateral fixation pattern in patients whose visual acuity could not be measured. Signs of low vision such as nystagmus, ocular bobbing or oculodigital massage were also noted. Orthoptic examination was performed with the prism cover test or Hirschberg and Krimsky tests depending on the cooperation level of the patient. After anterior segment examination, cycloplegic retinoscopy was carried out in all patients to find out refractive errors, and confirmations were provided via autorefractometry (Retinomax; Right Manufacturing, Virginia Beach, Virginia, USA) if possible. Posterior segment examination was performed with an indirect ophthalmoscope and with ultrasonography when required.

Cases with a significant visual impairment, which was defined as BCVA less than 0.3 in their better eye using the Snellen chart in accordance with WHO criteria, were included in the present study. Cases who had no central fixation and who had signs of very poor vision, like searching nystagmus, ocular bobbing or oculodigital massage, were also included. For the purpose of the study, major pathology that had the greatest effect on vision was taken into consideration in cases with multiple ocular pathologies. Bilateral decreased visual response caused by damage to either the visual cortex or the geniculostriate visual pathways based on MRI findings was defined as cortical visual impairment (CVI), unless any ocular abnormalities related to visual loss existed. The results of ophthalmologic examinations and coexistence of systemic diseases were documented in 695 patients (5.9%) who met the inclusion criteria of the present study, and were compared with our previously published clinic data on the aetiology of childhood blindness in Izmir before 2002.8

The data were stored on a computerised database and analysed using SPSS V.16.0 for Windows (SPSS, Chicago, Illinois, USA). Pearson χ2 and Yates corrected χ2 tests were used for statistical analysis and a p value of below 0.05 was considered as significant.


A total of 695 patients (5.9%) were enrolled in the present study, of whom 292 (42.0%) were female and 403 (58.0%) were male. No specific reason for the male preponderance was present in our hospital-based study. The mean age of patients at the first ophthalmologic examination in our Paediatric Ophthalmology and Strabismus Unit was 47.0±51.9 months, ranging from 1 month to 17 years (median: 24 months). The majority of the study participants were observed before the age of 47 months (mean age of the study population), accounting for almost two-thirds of all cases. Patients who presented before 1 year of age constituted 44.3% of the entire study population. Visual acuity was assessed using the Snellen chart in 274 cases (39.4%), and with Teller acuity cards in 77 cases (11.1%), whereas visual perception could not be measured in 344 patients (49.5%) in whom searching nystagmus, loss of fixation, ocular bobbing, or oculodigital massage were present due to the major ocular or neurologic pathologies that cause SVI. Nystagmus was found in 197 patients (28.3%), most frequently of the pendular type (14.2%; 99 cases). Strabismus was present in 200 patients (28.8%); among cases with horizontal strabismus, esotropia was more commonly seen in cases with a congenital aetiologic factor (55.5%), although a frequent type of strabismus was exotropia in patients with poor vision related to acquired aetiologic factor (58.4%). Exotropia was more frequently seen in patients with CVI (p<0.001). Cases with retinal or lenticular disorder had a tendency to develop eso deviations (p=0.019 and p=0.045, respectively). The mean angle of strabismus was found to be 22.9±12.0 prism dioptres (8–75 PD) and 18.5±8.4 prism dioptres (8–45 PD) in patients with eso and exo deviations, respectively. Developmental delay defined as a problem in one or many areas such as gross and fine motor skills, social and emotional skills, language and cognitive skills that was diagnosed by a paediatric neurologist was present in 16.8% of the study population. Demographics, ophthalmologic findings and concomitant systemic pathologies are shown in table 1.

Table 1

Demographics, ophthalmologic findings, and concomitant systemic pathologies

Blindness by anatomic site

Two hundred and twelve patients (30.5%) had CVI and 20.3% of those had a history of premature birth. Asphyxia, metabolic disorders, genetic syndromes, infections and trauma were identified as the other risk factors for CVI in our study population. The following three most frequently involved anatomic sites were retina (24.6%), crystalline lens (17.1%) and optic nerve (9.6%), which are summarised in table 2.

Table 2

Anatomic site of abnormality leading to visual impairment

Aetiology according to the time of insult leading to visual impairment

Genetic disorders were identified in 26.2% of patients. Chorioretinal dystrophies were found in 64 cases (9.2%). Neurometabolic disorders including mucopolysaccharidosis, galactosemia, urea cycle disorders, Lowe syndrome, Rett syndrome and Canavan disease were diagnosed in 33 cases (4.7%). Oculocutaneous albinism (3.9%), Down's syndrome (2.4%), bilateral retinoblastoma (0.9%) and Leber congenital amaurosis (0.7%) were the other common genetic abnormalities in our study population. The leading disorders within the prenatal/perinatal period were retinopathy of prematurity (ROP) (6.9%), congenital cataract (6.3%), cerebral pathologies such as lissencephaly, microcephaly, hypoplasia or agenesis of the corpus callosum (3.9%), congenital glaucoma (2.9%), optic nerve disorders (2.7%), and globe pathologies (2.1%). Frequently diagnosed disorders related to the postnatal period were secondary optic atrophy (3.2%), neoplasm (1.6%), meningitis (1.0%) and traumatic brain injury (0.9%). The time of insult was unknown in 278 patients (40.0%), of whom 61.5% had CVI, including hypoxic ischaemic encephalopathy (HIE), cerebral palsy, neurometabolic disorders and hydrocephalus. Chorioretinal dysplasia (21.6%) and primary optic atrophy (9.0%) were the other common pathologies in such cases.

Avoidable causes

In the present study, the leading pathology that could be encountered in the treatable causes of SVI was congenital and juvenile cataracts (16.4%). A family history of consanguinity was found to be present in 196 of the total 695 cases (28.2%) in our study population (table 3). Frequency of consangineous marriage was higher, especially in cases with congenital glaucoma (45.0%), optic nerve disorders (40.3%) and globe pathologies (37.5%). In 136 cases (19.6%), SVI was found to be related to various pathologies, which mostly may be preventable with genetic counselling such as Down's syndrome, metabolic disorders, retinoblastoma and hereditary chorioretinal dystrophies. Infectious causes primarily related to SVI were diagnosed in 1.2% of the cases (table 3).

Table 3

Avoidable causes of visual impairment

Comparison with our previous clinic data

When we compared the current results with our previously published clinic data concerning the aetiology of childhood blindness in Izmir before 2002, a statistically significant increase in the incidence of CVI (p=0.046), and a decrease in the incidence of uveal disorders (p<0.001) were found. Prevalence of blindness secondary to ROP reduced by a third (p=0.280), and a significant decrease was also achieved in aphakia-related SVI (p=0.028) over the last decade (table 4).

Table 4

The comparison of current results with our previously published clinic data8


As it is estimated that 500 000 children will become blind each year, the emotional, social and economic costs of childhood blindness to society have been gradually rising.1 ,2 Early diagnosis helps to reduce childhood mortality since the causes of blindness may also be related to lethal diseases, especially in the early decades of life.2–6 Up to 60% of newly diagnosed blind children die within 2 years of becoming blind in rural communities, however the UK surveillance study reported a figure as low as <10%.2–4 The prevalence of childhood blindness varies from 0.02% to 0.12% according to the socioeconomic status of the studied population, and over 70% of children suffering from blindness worldwide live in the poorest countries of Africa and Asia.4–6

Avoidable causes include preventable and treatable disorders that are related to childhood blindness. Avoidable measures such as refractive errors, keratitis, conjunctivitis and ocular trauma have been reported as the leading causes of childhood eye morbidity and SVI in rural communities.9–14 However, in developed countries, unavoidable causes including genetic diseases, congenital anomalies and disorders of the central nervous system have been reported as the main aetiologic factors for SVI.1 ,6 ,15 The frequency of unavoidable causes varies in different case series between 25.5% and 82% according to the socioeconomic status of the communities. Kong et al6 reported that a mean of 51% of SVI cases were found to be related to avoidable causes in developing countries. In the present study, the prevalence of diagnosis of any unavoidable aetiologic factors was found to be 38.3%, which is in accordance with the data of developing countries. A comparison of causative aetiologic factors for SVI between our study and other selected studies is given in table 5.

Table 5

The comparison of causative aetiologic factors for SVI between our study and other selected studies

Malnutrition and especially vitamin A deficiency as well as infectious diseases of the ocular surface or adnexa such as trachoma have still been encountered and are considerable causative factors of childhood blindness in low-income countries.9–14 A lack of such disorders within the aetiologic factors of SVI in our study population may be attributed to the national health programme, which has improved over the past four decades. In addition, our tertiary referral centre that is located in the western region of the country and covers a hinterland with comparatively higher socioeconomic and sociocultural status may also have a positive impact on such data. Since approximately 65% of the entire study population were referred to our clinic before the age of 4 years, blindness secondary to infections of the cornea or conjunctiva was not seen in any of the cases due to early diagnosis and complete treatment of keratitis or severe conjunctivitis, which could have progressed to corneal scarring without intervention. Correction of refractive errors with appropriate spectacles, setting an amblyopia treatment with eye patching if required, and scheduling regular follow-up visits in the first decade of life have helped to eliminate refractive amblyopia in many study cases. The low prevalence of refractive amblyopia in the present study should also be attributed to the improving performance of school health services that regularly monitor children for refractive errors and refer affected subjects to an ophthalmologist.

Compared with our published clinic data, a statistically significant decrease was detected in the incidence of preventable causes and especially genetic disorders. In 2002, the rate of consanguineous marriage was reported to be 15.1% among subjects aged 15 years or older, and 55.8% of those married their first-degree relatives.16 A population-based survey was conducted in 2011 on young adults between the ages of 15 and 24 years that represented 16.8% of the entire Turkish population. The frequency of consanguineous marriage was found to be 21.2% in this survey; furthermore, participants who married a first-degree relative constituted 51.9% of the endogamous population.17 Owing to improvements in invasive care facilities, the survival rate of children with various genetic or metabolic disorders has increased. Cases with perinatal asphyxia and related HIE that are significant risk factors for several mental and neurological disorders have also become more likely to survive during recent decades. In accordance with the present study, CVI has been found to be the most frequent aetiologic factor for childhood blindness in the USA.6 Khetpal and Donahue also reported that premature birth with a prevalence of 29% was one of the major risk factors for CVI.15 In our study, CVI was present in 212 cases (30.5%) and 20.3% of those were born prematurely. Any stage of ROP coexisted in 35 cases with CVI (16.5%) in the present study, and none of them developed end-stage retinopathy.

Retina is the most common anatomic site of involvement in childhood blindness worldwide.4 The incidence of ROP has increased in the last decades because of the augmented frequency of premature births relevant to the developments in assisted reproduction techniques, and advances in neonatology that allow a great improvement in survival rates of more immature neonates. However, we found a reduction of approximately one-third in the frequency of blindness secondary to ROP in the last decade. With the institution of carefully timed screening and scheduling proper treatment with laser photocoagulation or vitreoretinal surgery in required cases, the frequency of ROP-related blindness has substantially decreased. Blencowe et al18 reported a retinopathy-related SVI rate of 10.8% among preterm babies who developed any stage of ROP in 2010. In the USA, ROP was one of the three leading causes of childhood blindness.6 In middle and low income countries, remarkably, a lower frequency of ROP-related SVI was reported in studies carried out on students in schools for the blind.9 ,14 ,19 ,20 Nevertheless such data are likely to be underrepresented due to higher mortality rates of premature infants in poor communities.

Among blind children, the frequency of lenticular disorders was found to be between 10% and 20% in developing countries and <5% in developed countries.1–7 ,14 ,20–22 Lenticular disorders that are still commonly related to TORCH infections have remained one of the main causes of SVI in low-income societies. Preventive general public health care policies such as maternal immunisation against rubella and rubeola have dramatically reduced the burden of blindness due to congenital cataract in such countries.4–7 ,11–14 ,19–24 In the present study, lenticular disorders were identified in 17.1% of cases with SVI, of whom 57.1% had aphakia following cataract surgery. Compared with our previous data, a significant decrease was found in the prevalence of aphakia-related SVI, however the frequency of aphakic glaucoma was not significantly different between two study populations. The considerable decrease in SVI in such cases may be due to an improvement in our clinical practice for visual rehabilitation, which included aphakic contact lenses in infancy, plus bifocals in toddlerhood, and intraocular lens implantation thereafter. Scheduling postoperative follow-up visits is also important to diagnose and treat any possible aphakia complications such as amblyopia, strabismus and glaucoma.

Because the nature of uveal disorders has been well understood and more potent anti-inflammatory therapeutics have been developed based on an immunomodulatory approach over the last two decades, the incidence of SVI secondary to uveal disorders has gradually decreased.25 Compared with our previous data, we also found a significant decrease in the frequency of SVI due to uveal disorders, which is in accordance with recent literature.

The potential limitation of this study is that the prevalence of SVI might have been underestimated as this is not a population-based study. However, with the exception of the article published by Limburg et al,22 all the studies discussed in table 5 were conducted as either hospital based or school for the blind based, and our present results were compared with our previously published work that was also a hospital-based study. Cetin et al8 reviewed the charts of 998 patients who attended our Paediatric Ophthalmology and Strabismus Unit between 1998 and 2002. In the present study, our sample size is approximately 10 times higher as we reviewed the charts of 11 871 patients who attended our Unit between 2002 and 2014. The significant difference in the sample size of these two papers was thought to be another limitation of this study. The present work was conducted in our tertiary referral centre, with admissions from a comparatively higher socioeconomic and sociocultural region of Turkey, so the results do not represent the exact rates for the distribution of SVI in the whole country. Another limitation of this study was that the major pathology with the greatest effect on vision was taken into consideration in cases with multiple ocular pathologies. Since it is not uncommon to find more than one coexisting factor for SVI development, deliberately attributing the SVI to only one factor might negatively affect the results.

Our data about the prevalence of SVI and causative aetiologic factors give a snapshot of the proportion of the population under 18 years of age who had suffered from any ocular disorder at any one point in time. In the present study, the frequency of SVI among children who were referred to an ophthalmologist was found to be 5.9%, which is approximately 10–20 times higher than the prevalence data obtained from true population-based studies. Governments ought to ameliorate their policies on developing national screening programmes for treatable causes, such as congenital cataract, amblyopia, ROP and congenital glaucoma. Genetic counselling and parental education are also crucial, especially in countries with a high rate of consanguinity.


Neither this manuscript nor one with substantially similar content under our authorship has been published or is being considered for publication elsewhere. We certify that any affiliations with or involvement in any organisation or entity with direct financial interest in the subject matter or materials discussed in the manuscript are disclosed in the paper.



  • Contributors I certify on behalf of myself and all coauthors that we have all participated sufficiently in the conception and design of this work, and interpretation of the data, as well as in writing the manuscript, to take responsibility for it and accept its conclusions.

  • Competing interests None declared.

  • Ethics approval Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas